[Issue]: GPU not found in Python for pytorch rocm

[Loungagna]

Problem Description

After successfully installing the pytorch rocm package (https://download.pytorch.org/whl/nightly/rocm6.2/torch-2.6.0.dev20241016%2Brocm6.2-cp312-cp312-linux_x86_64.whl), any test of torch.cuda.is_available() fails.

rocminfo works fine, all tests from the CLI works fine for ROCM and HIP. python version is set via mise.

Operating System

24.04.1 LTS (Noble Numbat)

CPU

AMD Ryzen 7 7840HS w/ Radeon 780M Graphics

GPU

Marketing Name: AMD Radeon Graphics Name: amdgcn-amd-amdhsa--gfx1102

ROCm Version

ROCm 6.2.2

ROCm Component

No response

Steps to Reproduce

`python3

import torch print(torch.cuda.is_available())`

gives False

(Optional for Linux users) Output of /opt/rocm/bin/rocminfo --support

ROCk module version 6.8.5 is loaded

HSA System Attributes

Runtime Version: 1.14 Runtime Ext Version: 1.6 System Timestamp Freq.: 1000.000000MHz Sig. Max Wait Duration: 18446744073709551615 (0xFFFFFFFFFFFFFFFF) (timestamp count) Machine Model: LARGE System Endianness: LITTLE Mwaitx: DISABLED DMAbuf Support: YES

========== HSA Agents

---

Agent 1

---

Name: AMD Ryzen 7 7840HS w/ Radeon 780M Graphics Uuid: CPU-XX Marketing Name: AMD Ryzen 7 7840HS w/ Radeon 780M Graphics Vendor Name: CPU Feature: None specified Profile: FULL_PROFILE Float Round Mode: NEAR Max Queue Number: 0(0x0) Queue Min Size: 0(0x0) Queue Max Size: 0(0x0) Queue Type: MULTI Node: 0 Device Type: CPU Cache Info: L1: 32768(0x8000) KB Chip ID: 0(0x0) ASIC Revision: 0(0x0) Cacheline Size: 64(0x40) Max Clock Freq. (MHz): 5137 BDFID: 0 Internal Node ID: 0 Compute Unit: 16 SIMDs per CU: 0 Shader Engines: 0 Shader Arrs. per Eng.: 0 WatchPts on Addr. Ranges:1 Memory Properties: Features: None Pool Info: Pool 1 Segment: GLOBAL; FLAGS: FINE GRAINED Size: 32147684(0x1ea88e4) KB Allocatable: TRUE Alloc Granule: 4KB Alloc Recommended Granule:4KB Alloc Alignment: 4KB Accessible by all: TRUE Pool 2 Segment: GLOBAL; FLAGS: KERNARG, FINE GRAINED Size: 32147684(0x1ea88e4) KB Allocatable: TRUE Alloc Granule: 4KB Alloc Recommended Granule:4KB Alloc Alignment: 4KB Accessible by all: TRUE Pool 3 Segment: GLOBAL; FLAGS: COARSE GRAINED Size: 32147684(0x1ea88e4) KB Allocatable: TRUE Alloc Granule: 4KB Alloc Recommended Granule:4KB Alloc Alignment: 4KB Accessible by all: TRUE ISA Info:

---

Agent 2

---

Name: gfx1102 Uuid: GPU-XX Marketing Name: AMD Radeon Graphics Vendor Name: AMD Feature: KERNEL_DISPATCH Profile: BASE_PROFILE Float Round Mode: NEAR Max Queue Number: 128(0x80) Queue Min Size: 64(0x40) Queue Max Size: 131072(0x20000) Queue Type: MULTI Node: 2 Device Type: GPU Cache Info: L1: 32(0x20) KB L2: 2048(0x800) KB Chip ID: 5567(0x15bf) ASIC Revision: 9(0x9) Cacheline Size: 64(0x40) Max Clock Freq. (MHz): 2700 BDFID: 50176 Internal Node ID: 2 Compute Unit: 12 SIMDs per CU: 2 Shader Engines: 1 Shader Arrs. per Eng.: 2 WatchPts on Addr. Ranges:4 Coherent Host Access: FALSE Memory Properties: APU Features: KERNEL_DISPATCH Fast F16 Operation: TRUE Wavefront Size: 32(0x20) Workgroup Max Size: 1024(0x400) Workgroup Max Size per Dimension: x 1024(0x400) y 1024(0x400) z 1024(0x400) Max Waves Per CU: 32(0x20) Max Work-item Per CU: 1024(0x400) Grid Max Size: 4294967295(0xffffffff) Grid Max Size per Dimension: x 4294967295(0xffffffff) y 4294967295(0xffffffff) z 4294967295(0xffffffff) Max fbarriers/Workgrp: 32 Packet Processor uCode:: 39 SDMA engine uCode:: 18 IOMMU Support:: None Pool Info: Pool 1 Segment: GLOBAL; FLAGS: COARSE GRAINED Size: 16073840(0xf54470) KB Allocatable: TRUE Alloc Granule: 4KB Alloc Recommended Granule:2048KB Alloc Alignment: 4KB Accessible by all: FALSE Pool 2 Segment: GLOBAL; FLAGS: EXTENDED FINE GRAINED Size: 16073840(0xf54470) KB Allocatable: TRUE Alloc Granule: 4KB Alloc Recommended Granule:2048KB Alloc Alignment: 4KB Accessible by all: FALSE Pool 3 Segment: GROUP Size: 64(0x40) KB Allocatable: FALSE Alloc Granule: 0KB Alloc Recommended Granule:0KB Alloc Alignment: 0KB Accessible by all: FALSE ISA Info: ISA 1 Name: amdgcn-amd-amdhsa--gfx1102 Machine Models: HSA_MACHINE_MODEL_LARGE Profiles: HSA_PROFILE_BASE Default Rounding Mode: NEAR Default Rounding Mode: NEAR Fast f16: TRUE Workgroup Max Size: 1024(0x400) Workgroup Max Size per Dimension: x 1024(0x400) y 1024(0x400) z 1024(0x400) Grid Max Size: 4294967295(0xffffffff) Grid Max Size per Dimension: x 4294967295(0xffffffff) y 4294967295(0xffffffff) z 4294967295(0xffffffff) FBarrier Max Size: 32 *** Done ***

Additional Information

If I need to submit this issue somewhere else, please let me know.

Thank you very much for all the effort.

[ppanchad-amd]

Hi @Loungagna. Internal ticket has been created to investigate your issue. Thanks!

[Loungagna]

BTW, I verified that I'm hitting rocm pytorch with:

>>> import torch
>>> print(torch.__version__)
2.6.0.dev20241016+rocm6.2

[harkgill-amd]

Hi @Loungagna, I was not able to reproduce this issue on both a baremetal Pytorch installation and using the rocm/pytorch docker image (Installation Steps). My setup involved a Navi33/RX 7600 dGPU which is a similar ISA as the Radeon 780M Graphics.

For reference, the torch version on my local install is 2.6.0.dev20241021+rocm6.2 and within the docker container is 2.3.0a0+gitd2f9472. Could you please give both of these methods a try and report back?

[taylding-amd]

Hi, this issue will be closed now due to inactivity. Please feel free to reopen it or ask follow-up questions should further assistance be required. Thanks!